What is a Good PFR in Poker? Bands by Seat

What PFR counts, and what it does not

PFR is the share of hands where you raised preflop at least once. The denominator is every hand dealt to you, including the ones you fold and the blinds you post, and the numerator counts the hand once however many times you raised it. Three things people expect it to count and it does not: posting a blind is not a raise, so a big blind that never acts voluntarily still lands in the denominator; a 3-bet is a raise, so your PFR includes re-raises as well as opens; and the size of the raise is irrelevant, so a min-raise and a pot-sized one are the same event. Because the denominator is every hand, PFR and VPIP are the two fastest stats on any HUD to become readable, which is why almost every read starts with the pair rather than with either number alone.

Where the band comes from

The 15-24% band is a population guideline, but it is not arbitrary, and you can check it against this site's own preflop ranges. Counting combinations of the 1,326 possible starting hands — which is what a tracker counts — the range library opens UTG 9.8%, MP 13.1%, CO 23.1%, BTN 40.0%, SB 29.7%. Now put six of those players at one table. The pot is still unopened when the button acts 60.3% of the time and when the small blind acts 36.2% of the time, so the share of all hands that turn into an open works out at 12.4%. Add the hands you raise because somebody else opened first — you face an open on 34.9% of hands, and the same library's 3-bet ranges re-raise 5.8% of those openers once each seat is weighted by how often it is the one who opened — and the model lands at 14.5%.

• •That is just under the band's floor of 15%, and the gap is the point: the model assumes nobody limps and nobody isolates.
• •Add the spots it leaves out and you climb into the band. Our iso-raise ranges attack a limper with 29.7% of hands from the button and 16.1% from the cutoff, and the button squeeze range is 5.1% of hands.
• •One counting note, because it explains an apparent contradiction on this site. The library describes its button range as "The widest profitable open.", which is a hand-class rank; 40.0% is the same range counted in combinations, which is what a tracker counts. Combinations give the lower figure, because each offsuit hand carries twelve of them and each pair only six.

What a figure outside the band means for your play

A single average hides the thing that matters, so read the seats before you read the total. A low PFR is nearly always a late-position problem: the button and cutoff are where nobody has acted yet and two blinds are sitting there, and a player at or below 15% is usually folding the bottom half of the button range. A high PFR is nearly always an early-position problem: opening under the gun at button frequencies means playing big pots from the worst seat against ranges that have you dominated. The other half of the read is the gap to your VPIP: with VPIP guideline 20-28% and PFR 15-24%, the healthy gap is 2-8 points. A gap of fifteen points with a normal-looking PFR means the extra hands are all calls, and that is the most reliable losing pattern in any database review.

How it differs by format

The band is measured on 6-max online micro-stakes cash, and it does not transfer everywhere. Full ring: the model above says the button alone turns 24.1% of the hands dealt to it into opens, more than any other seat, and in a nine-handed game you sit there one hand in nine instead of one in six. Same ranges, lower PFR, and there is no separate full-ring band in the registry to quote you. Tournaments: stack depth changes the ranges themselves rather than your discipline in applying them, and under roughly 15 big blinds the push/fold chart replaces opening ranges altogether, so this band stops meaning anything. Live: a limp-heavy table hands you iso-raise spots instead of open-raise spots, which is a different mix of raises rather than more of them. Where the evidence is a band for one format, quoting it for another is guesswork with a decimal point on it.

How long before your own PFR means anything

Every hand dealt is a PFR opportunity, which makes this one of the quickest numbers to trust. Take the band's own width as the margin you need — it is 9 points wide, so 4.5 points either side of the middle — and the ordinary standard error of a proportion puts that at about 300 hands. To pin it to two points, so that the middle of the band and its edge are distinguishable, takes about 1,500 hands. The positional breakdown is the expensive part: each seat gets a sixth of your hands, so a readable PFR for each of six seats needs roughly six times the sample. That is the honest reason your first thousand hands can only tell you whether you are in the right neighbourhood.

Check your own figure

Our free leak report reads a hand history you already have, computes PFR alongside the rest of the preflop stats, and marks each one against these bands rather than against a number somebody typed into an article. It also splits the total by seat, which is where the actual answer lives. The range library holds the ranges the band was checked against.